Struct SanitizerHook 

pub struct SanitizerHook { /* private fields */ }

Post-invocation hook that runs the OutputSanitizer over tool results.

Behavior:

• If no sensitive data is detected, returns Allow.
• Otherwise, returns Redact(sanitized) with a JSON value whose strings have been sanitized in place (structure preserved).
• Emits GuardEvidence summarizing the findings so they flow into the kernel’s receipt. Raw secrets are never included; only previews, spans, and detector ids.

Implementations

impl SanitizerHook

pub fn new() -> Self

Build a sanitizer hook with the default sanitizer configuration.

pub fn with_config( config: OutputSanitizerConfig, ) -> Result<Self, OutputSanitizerConfigError>

Build a sanitizer hook with a custom sanitizer configuration.

pub fn from_sanitizer(sanitizer: OutputSanitizer) -> Self

Build a sanitizer hook from a pre-constructed sanitizer.

pub fn with_name(self, name: impl Into<String>) -> Self

Override the hook name (useful for telemetry).

pub fn sanitizer(&self) -> &OutputSanitizer

Access the underlying sanitizer (useful for tests / operator tooling).

Trait Implementations

impl Default for SanitizerHook

fn default() -> Self

Returns the “default value” for a type.

impl PostInvocationHook for SanitizerHook

fn name(&self) -> &str

fn inspect( &self, _ctx: &PostInvocationContext<'_>, response: &Value, ) -> PostInvocationVerdict

fn take_evidence(&self) -> Option<GuardEvidence>